Head motion during brain imaging with PET or other imaging modalities reduces image quality and compromises quantification of phenomena revealed in these images. Studies were undertaken during this period to (1) characterize the kind and amount of head motion occurring in normal subjects during simulated brain imaging, (2) to compare these normal results with motion measurements obtained in patients in whom cooperation could not be assumed, (3) to determine the extent to which these motions were reduced in normals and patients by a custom molded face mask created for each subject, and (4) to develop and validate analysis software that permitted these motion measurements obtained with an electromagnetic spatial tracking device to be interpreted. Based on the results obtained in these studies, head supports and associated apparatus for utilizing the thermoplastic mask technique for head restraint have been fabricated, or are now being fabricated, for all PET scanners in the DNM. Experimental studies will continue to be performed to determine whether further reductions in head motion, particularly chin-up/chin-down motions (or head motion about an axis through the ears) can be achieved with this (or another) restraint method.